CONTAGION (BBC4, 2018): The mathematical modeling of an outbreak

[Bill Ryan]

This is very, very interesting. I'm no longer a UK resident, so I was unaware of this until a few days ago.

In the VERY well-made and interesting 75 minute documentary, presenter Hannah Fry (super-friendly and pleasant, and herself a math PhD) volunteers as an imaginary Patient Zero, having just returned from an overseas vacation with a virus she doesn't know she has.

She then wanders around a small town doing very normal things, but periodically "infects" others — the potential victims all equipped with a special smartphone app to track their movements and behavior. The experiment used 10,000 nationwide volunteers.

All this is virtual, of course. The virus is defined to be potentially infectious within a range of 20 meters, but the probability of catching it increases with (a) proximity and (b) length of exposure. And there's no natural immunity to it.

The program then tracks how very quickly and easily the imaginary disease (conceived as a kind of flu) spreads to every corner of the British Isles. It also goes on to discuss what the public health response might or should be.

Irrespective of one's personal views about public health responses now in real time, this is how viruses work. As Chris Martenson has often reminded us, no virus ever cares what your opinions may be. They'll do their thing anyway. (And yes, viruses are real. )

Do watch. It's totally mainstream, of course, and basically recommends personal tracking and vaccination, with well-presented mathematical justification.

It shows that even without any Machiavellian agendas of any kind, this is the sort of modeling that drives administrators' decisions about vaccinations, home quarantine, travel restrictions, and the closure of schools and public gathering places.

But it's very interesting even just to understand some of how this modeling is used. (Math, like a virus, doesn't care about your views either. Math is just math. ) And as one would expect from the BBC, it's extremely well done. Eerily prescient, too. Recommended.

Source: https://youtube.com/watch?v=RmGiDUczhqQ

[Bill Ryan]

The video was posted on YouTube in April this year (2020). The comments are worth reading. A great many of them are along the lines of:

"So if they knew so much about this two years ago, how come the current response was so shambolic?"

[EFO]

Posted by Bill Ryan (here)
This is very, very interesting. I'm no longer a UK resident, so I was unaware of this until a few days ago.

In the VERY well-made and interesting 75 minute documentary, presenter Hannah Fry (super-friendly and pleasant, and herself a math PhD) volunteers as an imaginary Patient Zero, having just returned from an overseas vacation with a virus she doesn't know she has.

She then wanders around a small town doing very normal things, but periodically "infects" others — the potential victims all equipped with a special smartphone app to track their movements and behavior. The experiment used 10,000 nationwide volunteers.

All this is virtual, of course. The virus is defined to be potentially infectious within a range of 20 meters, but the probability of catching it increases with (a) proximity and (b) length of exposure. And there's no natural immunity to it.

The program then tracks how very quickly and easily the imaginary disease (conceived as a kind of flu) spreads to every corner of the British Isles. It also goes on to discuss what the public health response might or should be.

Irrespective of one's personal views about public health responses now in real time, this is how viruses work. As Chris Martenson has often reminded us, no virus ever cares what your opinions may be. They'll do their thing anyway. (And yes, viruses are real. )

Do watch. It's totally mainstream, of course, and basically recommends personal tracking and vaccination, with well-presented mathematical justification.

It shows that even without any Machiavellian agendas of any kind, this is the sort of modeling that drives administrators' decisions about vaccinations, home quarantine, travel restrictions, and the closure of schools and public gathering places.

But it's very interesting even just to understand some of how this modeling is used. (Math, like a virus, doesn't care about your views either. Math is just math. ) And as one would expect from the BBC, it's extremely well done. Eerily prescient, too. Recommended.

Source: https://youtube.com/watch?v=RmGiDUczhqQ

The same here with London riot...
I'm still wondering what if all these "mathematical prediction" are applied when all "equation's elements" give the "expected" result?At least for me it's obvious...it's happened so many times now.

re: publica 2014 - Hannah Fry: I predict a riot!
(1:03:52 min.)

Source: https://youtube.com/watch?v=ROnjZDdt8O8

[Bill Ryan]

Posted by EFO (here)
The same here with London riot...
I'm still wondering what if all these "mathematical prediction" are applied when all "equation's elements" give the "expected" result?At least for me it's obvious...it's happened so many times now.

re: publica 2014 - Hannah Fry: I predict a riot!
(1:03:52 min.)

Source: https://youtube.com/watch?v=ROnjZDdt8O8

This is really about how the laws of probability apply to large numbers of individuals.

Many years ago, I read an article by a mathematician who was seriously questioning the laws of probability as we understood them. (These are basically unchanged since the work of Blaise Pascal in 1654.)

He made this argument. I don't have the original, but the numbers were very similar to this. It's all about the numbers of New York postmen who were bitten by dogs each year.

• 1980: 72 postmen were bitten by dogs.
• 1981: 69 postmen were bitten by dogs.
• 1982: 71 postmen were bitten by dogs.
• 1983: 70 postmen were bitten by dogs.
• 1984: 73 postmen were bitten by dogs.
• 1985: 70 postmen were bitten by dogs.
• 1986: 71 postmen were bitten by dogs.

The number was almost exactly the same every year. His utterly serious question was:

How do the dogs know when to stop biting the postmen?

[silvanelf]

Posted by Bill Ryan (here)
This is really about how the laws of probability apply to large numbers of individuals.

Many years ago, I read an article by a mathematician who was seriously questioning the laws of probability as we understood them. (These are basically unchanged since the work of Blaise Pascal in 1654.)

He made this argument. I don't have the original, but the numbers were very similar to this. It's all about the numbers of New York postmen who were bitten by dogs each year.

• 1980: 72 postmen were bitten by dogs.
• 1981: 69 postmen were bitten by dogs.
• 1982: 71 postmen were bitten by dogs.
• 1983: 70 postmen were bitten by dogs.
• 1984: 73 postmen were bitten by dogs.
• 1985: 70 postmen were bitten by dogs.
• 1986: 71 postmen were bitten by dogs.

The number was almost exactly the same every year. His utterly serious question was:

How do the dogs know when to stop biting the postmen?

The dogs know about the the laws of probability.

[EFO]

Posted by Bill Ryan (here)
His utterly serious question was:

How do the dogs know when to stop biting the postmen?

When the postmen are replaced?

For me the laws of probability sound like I would want to measure an incertitude associated with a random variable...Sheesh

[Satori]

When I was in college in the mid 1970's I took a class on statistics. Very interesting class. The book I remember the most, which was one of several we had to buy for the class, was called: "How To Lie With Statistics." It gives the reader insight into probability theory and how to manipulate not just the data (i. e., "facts"), but the theories and tools of statistics, to achieve the outcome you want (on paper at least), within certain parameters without technically lying--or at least not making your self liable for a clear case of fraud.

Edit: I should add that the purpose of the book is not to enable one to commit fraud or lie with statistics, but rather to detect it when someone is massaging the data.

The bottom line is that, and this is true in all scientific and technical fields of endeavor, one expert on this side of an issue can come out and say “x” and another expert on that side of the issue come out and say “not x”, and it is extremely difficult to establish who is correct and to know who to believe. Hence, there has been a push in the past several decades to exalt experts, so called, and place them on a pedestal, to try to convince the public to go with your expert and not the other side’s expert.

But, both “experts” could be wrong. All “experts” could be wrong.

Experts have to be qualified to give an opinion and their opinions have to be both relevant and reliable.

[Satori]

I’m bumping my post, immediately above, because I added some stuff.

[araucaria]

The best available research presented in this video comes up with the overblown statistics that the scientist who had to resign for flouting lockdown rules [edit: Neil Ferguson] had been churning out for twenty years and is the UK government’s excuse for still being smug with over 65,000 surplus deaths. So what went wrong? The phone app seems to work fine; the problem has to be to do with the parameters used for producing and analyzing the data. The R number at 1.8 sounds fair enough, but how it is applied to the infection rate may be problematic (I am a layman in these subjects). Infections seem to be defined in terms of proximity and duration: come too close for too long and the outcome is automatic. If so, then we have a problem similar to the issue of the number of New York postmen bitten by dogs in any one year: how do the dogs know when to stop? How do people know how much social distancing to do in order to reach exactly 1.8? One of the two parameters must be overriding the other. (I think New Yorkers probably have a constant number of aggressive dogs. If those dogs were identified put down, and not replaced, you could conceivably get the bite rate down to zero.)

The other problem then is to do with the infection rate itself. The relatively low real-life Covid figures suggest that the proximity/duration levels as defined are too sensitive: people generally are not that contagious and not that vulnerable. A tiny error at the outset will produce a huge error in the total figure. The error is compounded here by the fact that the sample population is explicitly stated to be younger than it should be in order to be representative. The old people in care homes who actually accounted for most of the deaths are hardly represented at all, and yet the calculated figures are an order of magnitude higher than real life. These are the people who don’t go to the shop or pubs or to London; since they don’t intermingle much, the R number in care homes must be extremely high, and much lower everywhere else.

So the answer to the question "So if they knew so much about this two years ago, how come the current response was so shambolic?", may well be: they didn’t know very much – a computer virus might work like this, but not a real one. We are told 3/4 of the population would be infected, or nearly one half if the superspreaders were vaccinated. (That would be if the vaccine were 100% reliable, when the actual figure is closer to 40%.) On that basis, any government would naturally be totally out of its depth, and herd immunity might well be the best response. Also any kind of lockdown begins to look like a very blunt tool. As things stand, no one has been vaccinated and still the figures are much lower. Following the science is a good idea, but the science needs to do better than this.

[onawah]

Hmmm, so that would be the BBC that Bill Gates donated millions to, yes? ....
No hidden agenda there!
See: https://youtube.com/watch?v=0W9C6k7neGw
and: More about the article from the Combia Journalism Review: https://21stcenturywire.com/2020/08/...fact-checkers/

Posted by Bill Ryan (here)
This is very, very interesting. I'm no longer a UK resident, so I was unaware of this until a few days ago.

In the VERY well-made and interesting 75 minute documentary, presenter Hannah Fry (super-friendly and pleasant, and herself a math PhD) volunteers as an imaginary Patient Zero, having just returned from an overseas vacation with a virus she doesn't know she has.

She then wanders around a small town doing very normal things, but periodically "infects" others — the potential victims all equipped with a special smartphone app to track their movements and behavior. The experiment used 10,000 nationwide volunteers.

All this is virtual, of course. The virus is defined to be potentially infectious within a range of 20 meters, but the probability of catching it increases with (a) proximity and (b) length of exposure. And there's no natural immunity to it.

The program then tracks how very quickly and easily the imaginary disease (conceived as a kind of flu) spreads to every corner of the British Isles. It also goes on to discuss what the public health response might or should be.

Irrespective of one's personal views about public health responses now in real time, this is how viruses work. As Chris Martenson has often reminded us, no virus ever cares what your opinions may be. They'll do their thing anyway. (And yes, viruses are real. )

Do watch. It's totally mainstream, of course, and basically recommends personal tracking and vaccination, with well-presented mathematical justification.

It shows that even without any Machiavellian agendas of any kind, this is the sort of modeling that drives administrators' decisions about vaccinations, home quarantine, travel restrictions, and the closure of schools and public gathering places.

But it's very interesting even just to understand some of how this modeling is used. (Math, like a virus, doesn't care about your views either. Math is just math. ) And as one would expect from the BBC, it's extremely well done. Eerily prescient, too. Recommended.

Source: https://youtube.com/watch?v=RmGiDUczhqQ